US4204251A - Interconnection unit for multiple data processing systems - Google Patents

Interconnection unit for multiple data processing systems Download PDF

Info

Publication number
US4204251A
US4204251A US05/865,116 US86511677A US4204251A US 4204251 A US4204251 A US 4204251A US 86511677 A US86511677 A US 86511677A US 4204251 A US4204251 A US 4204251A
Authority
US
United States
Prior art keywords
control
data processing
signal
information
processing system
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
US05/865,116
Other languages
English (en)
Inventor
Finn Brudevold
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
ATEX PUBLISHING SYSTEMS Corp
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to US05/865,116 priority Critical patent/US4204251A/en
Priority to GB7850070A priority patent/GB2011680B/en
Priority to GB8139088A priority patent/GB2091917B/en
Priority to JP16459178A priority patent/JPS54102932A/ja
Priority to DE19782856483 priority patent/DE2856483A1/de
Application granted granted Critical
Publication of US4204251A publication Critical patent/US4204251A/en
Assigned to ATEX, INC. A DE CORPORATION reassignment ATEX, INC. A DE CORPORATION ASSIGNMENT OF ASSIGNORS INTEREST. Assignors: BRUDEVOLD, FINN
Assigned to ATEX PUBLISHING SYSTEMS CORPORATION reassignment ATEX PUBLISHING SYSTEMS CORPORATION ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: ATEX INC.
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F15/00—Digital computers in general; Data processing equipment in general
    • G06F15/16—Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
    • G06F15/163—Interprocessor communication
    • G06F15/167—Interprocessor communication using a common memory, e.g. mailbox

Definitions

  • This invention generally relates to data processing systems and more specifically to the interconnection of data processing systems in a distributed data processing network.
  • a "file” is an organized collection of information directed to some purpose.
  • a file may comprise a set of program instructions for use by a data processing system or a block of data to be processed in response to instructions stored in some other file.
  • a file is specified by a unique identifier. Typically, this identifier is a common name which the person forming the file assigns to it.
  • a "distributed data processing network” includes interconnected data processing systems. Such a network is characterized in that each data processing system operates independently of the other systems and yet has the capacity for transferring files with any other system in the network.
  • Data processing systems in the past have been interconnected to operate as distributed data processing networks.
  • two data processing systems are interconnected by means of a linking circuit.
  • This type of interconnection enables various devices to connect to each data processing system and operate independently with that system while the link transfers files between the systems.
  • the files are stored redundantly.
  • Such distributed data processing networks are called replicated networks. They are relatively simple to implement when two data processing systems are interconnected. However, if one attempts to construct a replicated network with three or more systems, the interconnection and the required control programming for operating the network becomes extraordinarily complex. Thus, replicated networks tend, from a practical standpoint, to be limited to networks that include only two independent data processing systems.
  • Another approach to interconnection of data processing systems in distributed data processing networks involves the interconnection of two or more data processing systems in a master-slave relationship.
  • one data processing system performs solely as a master system. All the other data processing systems connect to the master system as slave systems. The master system controls any communications between the individual slave systems.
  • the master system When the master system is a dedicated data processing system, however, much of the redundancy in the replicated system is lost. As will be apparent, if the master system malfunctions and ceases to operate, the entire network becomes inoperative. In another master-slave implementation, all the data processing systems are in parallel, but one system is additionally programmed to operate as the master system. This reduces the susceptibility to malfunctions because the master program may be transferred to any other data processing system in the network. However, the combined load on a data processing system additionally operating as the master system can overload the master system. Any such overloading that affects the operation of the master system slows the overall efficiency of the entire network.
  • control programs As the control programs become more complex in all the foregoing approaches, they require more time to process. However, these control programs perform overhead operations that require time that therefor can not be allocated to processing programs that produce useful output. Increasing the time for processing these control programs therefore makes the data processing systems run significantly slower; that is, the processing time for programs in the distributed data processing network is greater than the time required if a single data processing system were to process the same program on a dedicated basis.
  • Another object of this invention is to provide apparatus for interconnecting multiple data processing systems that comprises electronic circuitry that affords great flexibility in the preparation of control programs.
  • Still another object of this invention is to provide apparatus for interconnecting multiple data processing systems that are simple to implement and to utilize.
  • Yet another object of this invention is to provide apparatus for interconnecting multiple data processing systems that minimizes processing time for control programs.
  • each data processing system in a distributed data processing network connects to an interconnection unit. If a particular data processing network requires a file to be transferred, it requests access to the interconnection unit. When that request is granted by the interconnection unit, exclusive control over the interconnection unit is transferred to the requesting data processing system. The requesting system then transfers information into a portion of a memory in the controller that can be accessed by all other data processing systems in the network. When the information is stored in the memory, the system with exclusive control can cause the interconnection unit to interrupt the operations of the system to which the information has been sent. In response the other system requests control over the interconnection unit thereby to retrieve and process the information.
  • FIG. 1 is a block diagram of a distributed data processing network constructed in accordance with this invention
  • FIGS. 2A through 2D disclose four specific registers that are associated with each interface circuit shown in FIG. 1;
  • FIG. 3 is a detailed block diagram of the control section of the controller and corresponding control section of the interface circuit shown in FIG. 1;
  • FIG. 4 is a detailed block diagram of the data section of the controller and corresponding data section of the interface circuit shown in FIG. 1;
  • FIG. 5 is a detailed block diagram of the bus control section of the interface circuit shown in FIG. 1;
  • FIG. 6 is a detailed block diagram of a host processor interface circuit shown in FIG. 1.
  • a distributed data processing network shown in FIG. 1 comprises three data processing systems designated by reference numerals 10, 11 and 12.
  • the system 12 represents data processing system "n" where "n” corresponds to the maximum number of data processing systems that can be connected in the network.
  • the specific constrution of each data processing system in the network depends upon the particular application of that system.
  • each data processing system will include, as a minimum, the units shown in the data processing system 10. These include a central processor unit 13 which communicates over a system bus 14 with a random access memory unit 15 and one or more peripheral units 16.
  • a peripheral unit may comprise input-output devices such as teletypewriters and video display-keyboard devices, output devices such as line printers and secondary storage devices such as disk memories.
  • the remaining data processing systems in FIG. 1 will have a similar structure.
  • an interconnection unit 17 constructed in accordance with this invention contains a number of circuits.
  • a host processor interface circuit connects to each data processing system in the network.
  • the host processor interface circuit 20 it connects to the system bus 14 and thereby interfaces the interconnection unit 17 with the data processing system 10.
  • Each of the host processor interface circuits 20, 21 and 22 also connects to a corresponding one of interface circuits 23, 24 and 25 by means of buses 30, 31 and 32.
  • Each interface circuit comprises several sections, and the interface circuit 23 is divided into those sections: namely, a bus control section 23B that connects to the bus 30, a control section 23C that connects to the bus control section 23B, and a data section 23D that also connects to the bus control section 23B.
  • Each of the control and data sections connects to a controller circuit 33.
  • the controller circuit 33 includes a control section 34, and a bus 35 connects the control section 34 to each interface circuit control section, such as control section 23C.
  • a data section 36 communicates with each interface circuit data section, such as the data section 23D, over a bus 37.
  • Each data processing system in the network shown in FIG. 1 operates independently of the other data processing systems. That is, each system responds to programs stored in its own memory and signals received from and transmitted to its own peripheral devices.
  • the interconnection unit 17 is connected to all the data processing systems. Therefore, it enables any one data processing system, as a "requesting system", to transfer a file to or from any other system, as a "target system", in the network.
  • a transfer is initiated when the requesting system requests control of either the control section 34 or the data section 36 in the interconnection unit 17.
  • the requesting system transfers the information to the corresponding one of the control and data sections 34 and 36.
  • the requesting system then can interrupt operations of the target system.
  • the data processing system 10 For example, if the data processing system 10, as part of its operation, needs to transfer a file to the data processing system 11, it is a requesting system and issues a request for control over the control section 34.
  • the host processor interface circuit 20 couples this request to the interface circuit 23 over the bus 30 and into the control section 23C.
  • the control section 34 of the controller 36 artibrates concurrent requests for control and recognizes one such request.
  • the controller circuit 33 grants control to the system 10
  • the control section 34 and control section 23C transfer a signal back to the host processor interface 20 to indicate to the data processing system that it has exclusive control over the interconnection unit 17.
  • the data processing system 10 specifies which data processing system, as a "target" system, is to receive the file and performs other control functions by transferring information to and from the control section 34, as the control section 34 contains information about each system in the network. Once all this information is transferred, the requesting system 10 relinquishes control of the control section 34 and requests exclusive control over the data section 36. An analogous process transfers exclusive control to the data processing system 10. Then the file to be transferred is routed through the host processor interface circuit 20, over the bus 30, through the bus control section 23B and data section 23D and over the bus 37 into the data section 36. Once the transfer is completed, the controller circuit 33 and interface circuit 24 interact to notify the target system, i.e., the data processing system 11, that the file is available.
  • target system i.e., the data processing system 11
  • the data processing system 11 then becomes a requesting system and obtains control of either the control section 34 or the data section 36 or both sections in sequence. Once control over the data section 36 had been obtained, the file is transferred from the data section 36 over the bus 37, through the interface circuit 24, over the bus 31 and through the host processor interface circuit 21 into the data processing system 11.
  • the data processing system 11 and the interconnection unit 17 could then interact to notify the data processing system 10 that the transfer had been completed.
  • the interconnection unit shown in FIG. 1 enables a requesting data processing system to gain exclusive control of and transfer information to the interconnection unit 17.
  • the interconnection unit 17 buffers that information until the target system is notified and, as a requesting system, completes the transfer.
  • the interconnection unit 17 thereby enables each system to transfer files with any other system in the network.
  • the "radial" nature of the interconnection of the systems is apparent from FIG. 1. This radial configuration also enables the transfer to occur between two data processing systems independently of any other system in the network.
  • the network shown in FIG. 1 is, according to the previous definition, a true distributed data processing network. Moreover, this result is achieved without any change in the structure of the individual data processing systems.
  • Each PDP11 data processing system communicates with the interconnection unit 17 by performing bus reading and writing operations over its system bus that corresponds to the system bus 14.
  • the central processor unit 13 becomes a "master" unit.
  • a master unit initiates a bus operation by first gaining control of the system bus 14. Then it transmits address and control signals onto the bus 14.
  • the address signals identify a specific storage location in the memory unit 15 or any peripheral unit 16; that unit is a "slave” unit for the duration of the bus operation.
  • Certain control signals called direction signals specify whether the bus operation is a reading or writing operation. If a writing operation is being performed, the master unit also transmits data onto the system bus 14 concurrently with the address and control information. Then the master unit transmits a master synchronizing signal.
  • the slave unit responds to these signals by performing the specified operation and then transmitting back to the master unit a slave synchronizing signal that enables the master and slave units to complete the bus operation.
  • the interconnection unit 17 contains a number of storage locations, or registers, that are accessible to each data processing system. Specifically, these locations comprise a set of four registers that are associated with each data processing system.
  • the register set associated with the data processing system 10 is shown in FIG. 2; the register sets associated with the other systems would be identical.
  • Each set includes a control status register 40 shown in FIG. 2A that contains various stages. These stages comprise circuitry in the control section 23C of the interface circuit 23 and are associated with the control section 34.
  • FIG. 2B depicts a control section data buffer 41.
  • FIG. 2C depicts a data section control status register 42 that comprises circuitry in the data section 23D of the interface circuit 23 while FIG. 2D depicts a data buffer 43 that also is a fictitious register that corresponds to a memory 43 in the data section that is shown in FIG. 4.
  • each control status register comprises a number of individual stages that are constituted by various flip-flops and gates in its respective interface unit control section.
  • the same reference numeral is used to identify the stage in FIG. 2 and to identify the specific circuit element in the other Figures that receives or transmits the corresponding signal.
  • the least significant stages of the control status register 40 are MY ID stages 50. These stages are read-only stages and reflect the signals transmitted by an ID SWITCHES circuit 50 shown in FIG. 3.
  • the circuit 50 produces a unique combination of MY ID signals for each data processing system. In this specific example, there are four bits. Therefore, this specific embodiment of the interconnection unit 17 can accommodate up to sixteen data processing systems.
  • a C INT (control system interruption) stage 51 is a write-only stage and is constituted by a NAND gate 51 shown in FIG. 3. Whenever the signal in the C INT bit position is asserted during a bus writing operation to the register 40, the NAND gate 51 is energized.
  • the interconnection unit 17 responds to the C INT signal by interrupting a data processing system that is specified by predetermined bit positions in CONTROL ADDRESS stages 52, these stages being constituted by an ADDRESS latch 52 in FIG. 3.
  • a flip-flop 53 in FIG. 3 constitutes a read-write C REQ (control section request) stage 53.
  • the interface circuit 23 establishes a request for control over the interconnection unit 17.
  • an AND gate 54 i.e., the read-only CB RDY stage
  • CB RDY control bus ready
  • a C IE (control section interruption enabling) stage 55 provides more efficient transfer of control to the data processing system.
  • a flip-flop 55 in FIG. 3 constitutes the read-write C IE stage. If the C IE flip-flop 55 is set when the CB RDY signal is asserted, the control section 23C and the host processor interface circuit 20 interrupt the data processing system 10. This provides a positive indication to the system 10 that it has exclusive control over the interconnection unit 17.
  • the stages in the ADDRESS stages 52 are read-write latch. In addition to identifying a data processing system when the C INT signal is asserted, these stages provide an address for a particular storage location in the memory 41 that will be accessed during a bus operation involving the control data buffer 41.
  • control section data buffer 41 is a fictitious register.
  • the interconnection unit 17 couples the location identified by the ADDRESS latch 52 to the data conductors in the bus 14.
  • an MPB CLR stage 60 is a write-only stage. When set, it initializes the interconnection unit 17. Corresponding circuitry is not shown in FIG. 4.
  • a FI CLR stage 61 also is a write-only stage shown as an AND gate 61 in FIG. 4. When the AND gate 61 is energized, a counter 62 in FIG. 4 is set to an initial value.
  • a P UP stage 63 is a ready-only stage that is set if the controller circuit 33 is activated. This stage thereby enables a data processing system to determine whether the interconnection unit 17 is turned on and operating. Its specific circuitry is not shown.
  • a W/R stage 64 is a read-write stage that comprises a flip-flop 64 in FIG. 4. This stage controls, in combination with the direction signals, the incrementing of the counter 62 during transfers to and from the memory 43.
  • a D INT stage 65 is a write-only stage represented by an AND gate 65 in FIG. 4.
  • AND gate 65 in one interface circuit When the AND gate 65 in one interface circuit is energized, another interface circuit connected to the data processing system identified in DESTINATION stages 66 interrupts that data processing system.
  • the stages 66 are represented as a DATA ADDRESS latch 66 in FIG. 4. These stages also coact with signals from the counter 62 to identify unique storage locations in the memory 43.
  • a DM REQ stage 67 corresponds to the C REQ stage 53 and is a read-write stage constituted by a flip-flop 67 in FIG. 4.
  • an AND gate 68 in FIG. 4 asserts a DM RDY signal. If the data processing system 10 then performs a bus reading operation with the control status register 42, it will ascertain that it has exclusive control over the control section 36 of the interconnection unit 17.
  • a read-write D IE stage 70 represented as a flip-flop 70
  • a number of conditions including the energization of the AND gate 68, can enable the host processor interface unit 20 to interrupt the data processing system 10.
  • notification of that control can be sent back to the system in a positive manner.
  • a write-only CLR INT stage 71 comprising AND gates 71A (FIG. 4) and 71B (FIG. 3) and a NAND gate 71C (FIG. 4) clears certain interruption conditions when it is energized.
  • a read-only DI PEN stage 72 comprises a flip-flop 72 in FIG. 4.
  • the flip-flop 72 is set by an interrupt command (i.e., the assertion of the D INT signal by another system) that also specifies the system 10.
  • the DI PEN signal indicates that this interruption has been received and is pending. If the D IE flip-flop 70 also is set, the data processing system 10 is interrupted. If the CLR INT and DI PEN signals are asserted during a bus writing operation to the control status register 42, the AND gate 71A clears the flip-flop 72.
  • a flip-flop 73 in FIG. 3 provides the signal for a read-only CI PEN stage 73.
  • the data processing system 10 is interrupted. If the CLR INT and CI PEN signals are asserted during a bus writing operation to the control status register 42, the AND gate 71B clears the flip-flop 73.
  • the memory 41 in FIG. 3 comprises a number of storage locations.
  • the CONTROL ADDRESS latch 52 contains eight stages so that any one of 256 storage locations can be specified.
  • sixteen storage locations in the memory 41 allocated each data processing system. These storage registers contain status and control information that is useful during the transfer of files. These storage locations might include, for example, information that indicates whether the memory 43 shown in FIG. 4 is full or empty. Other information might include the identification of a requesting data processing system.
  • the memory 43 comprises blocks of locations which correspond to a convenient data block size.
  • files are stored in disk memory units.
  • the basic block of data in such disk memory units is a sector having 256 words. Therefore, the memory 43 comprises 4096 storage locations, and 256 storage locations are assigned to each data processing system.
  • FIG. 5 depicts data paths through the bus control section 23B of the interface circuit 23 shown in FIG. 1. All the interface circuits have the same data paths.
  • signals are transferred between the host processor interface circuits and corresponding interface circuits by using differential signal transfers.
  • incoming data signals on the bus 30 energize differential receivers 44.
  • BD (bus data) signals from the receivers 44 pass through either a transceiver circuit 45 as CBD (control bus data) signals to the data section 35D of the bus 35 or a transceiver circuit 46 as DBD (data bus data) signals to the data section 37D of the bus 37.
  • CBD control bus data
  • DBD data bus data signals
  • incoming CBD and DBD signals pass through the respective transceivers 45 and 46 to a four-input multiplexer 47 as CD (control section data) and DD (data section data) signals respectively.
  • the other multiplexer inputs receive CCSR (control section control status register) and DCSR (data section control status register) signals respectively.
  • Address (A1 and A2) signals select which multiplexer input signals will be coupled through the multiplexer 47 to differential transmitters when an MUXEN signal is asserted.
  • the differential transmitters 48 transmit the selected signals onto the bus 30 in response to a DREN (data read enable) signal.
  • the data processing system 10 initiates the transfer by obtaining control of control and data sections 34 and 36 in sequence.
  • control and data sections are independent.
  • independent I mean that the two sections operate separately, so it is possible for the control section to be under exclusive control of one data processing system while the data section is under the exclusive control of another data processing system.
  • a file transfer is initiated when the requesting data processing system 10 performs a bus writing operation to the control status register 40 in the control section 23C of the interface circuit 23.
  • the C INT stage 51 and the CONTROL ADDRESS stages 52 are not be used.
  • the C REQ stage 53 is set.
  • the C IE stage 54 also is set.
  • the central processor unit 13 When the central processor unit 13 initiates the bus writing operation, it asserts a BC1 signal indicating a writing operation and a master synchronizing (BMSYN) signal in addition to the data and the address for the control status register 40.
  • the address signals are conveyed to an address comparator 80 over address conductors 14A and control conductors 14C in the bus 14.
  • the master synchronizing (BMSYN) signal enables the address comparator 80 to produce an output signal if the incoming address signals on conductors 14A correspond to any of the predetermined addresses for the registers shown in FIG. 2. If one of the four addresses is received, the address comparator 80 and the BMYSN signal energize an AND gate 81 thereby to transmit an ME signal.
  • the ME signal is conveyed through a delay circuit 82 to differential transmitters 83 to appear as MS signals on the bus 30.
  • the ME signal also is conveyed to an AND gate 84 that is energized by the asserted BC1 signal during a writing operation to transmit a DATO signal.
  • the DATO signal from the AND gate 84 during each bus writing operation conditions differential transmitters 85 to couple signals from the data conductors 14D in the bus 14 onto data conductors in the bus 30.
  • the DATO signal also controls, through an inverter 86 and an AND gate 87, the transfer of data signals through driving circuits in data transceivers 88 during bus reading operations.
  • the differential transmitters 83 also convey low-order address signals A1 and A2 to identify one of the four registers and a C1 signal, derived from the BC1 signal, onto corresponding conductors in the bus 30 during bus reading and writing operations.
  • differential receivers 89A pass the A1 and A2 address signals to a decoder 90.
  • Differential signals produced by the MS signals pass through a differential receiver 91 to form the BMSYN signal.
  • an AND gate 92 transmits a D-TO-M control signal. This signal conditions the transceivers 45 and 46 to convey the signals from the differential receivers 44 onto either the CBD conductors 35D or the DBD conductors 37D if the CB RDY or the DM RDY signal is asserted.
  • the BMSYN signal controls the output signal from the decoder 90.
  • an AND gate 93 produces a writing pulse (WP). More specifically, the BC1 signal from the differential receivers 89 and the BMSYN signal from the differential receivers 91 energize the AND gate 93 until a delay circuit 94 disables the AND gate 93 through an inverter 95.
  • the resulting writing pulse enables the decoder 90 to produce either an LCCSR or LDCSR pulse depending upon the address signals from the differential receivers 89A. In this case the decoder 90 transmits an LCCSR pulse that loads the various read-write and write-only stages of the control status register 40.
  • the signal from the delay circuit 94 and the A1 signal that specifies the control status register 40 energize an AND gate 96 that produces a CSRSSYN signal.
  • This signal is coupled through OR gates 97 and 98 and through an AND gate 99 that is enabled by the BMSYN signal.
  • the AND gate 99 energizes differential transmitters 100 thereby to produce SS signals that are conveyed back to the host processor interface circuit 20 in FIG. 6.
  • Differential receivers 101 couple the SS signals onto the bus 14 as a BSSYN signal. As known, the BMSYN signal then will terminate and the AND gate 99 in FIG. 5 will then terminate the SS signals. This completes the bus writing operation. Once this bus writing operation is completed, the data processing system can begin to process some other program.
  • a clock 102 produces T1 and T2 timing pulses that are displaced in phase.
  • Each T1 pulse is applied to a clocking (CLK) input of a counter 103 which has a modulus corresponding to the number of data processing systems connected in the data processing network. In this specific embodiment the modulus is 16.
  • Comparators in each interface circuit receive the output signals from the counter 103.
  • the comparator 104 energizes an AND gate 105 that is enabled when the flip-flop 53 is set.
  • the AND gate 105 transmits a C REQ signal onto a corresponding one of the conductors 35C.
  • This C REQ signal conditions a C GRANT flip-flop 106 in the control section 34 to be set by the next T2 pulse from the clock 102.
  • this flip-flop sets, it generates a C GRANT signal and disables the counter 103 so the counter 103 can not increment to any successive number.
  • the C GRANT signal is coupled back to all the interface circuits over a corresponding one of the conductors 35C and energizes the AND gate 54 in the interface circuit 23 when that gate is enabled by the flip-flop 53 and the comparator 104 thereby to transmit the CB RDY signal.
  • the assertion of the CB RDY signal therefore indicates that exclusive control over the control section 34 has been assigned to the data processing system 10.
  • the CB RDY signal energizes an AND gate 107 that is shown in both FIGS. 3 and 5. As shown in FIG. 5, the AND gate 107 energizes differential transmitters 108 to transmit CG signals onto the bus 30. Differential receivers 101 at the host processor interface circuit 20 in FIG. 6 produce a CGOINT signal in response to the CG signals.
  • the CGOINT signal is one input to a dual interruption circuit 109 that responds to the CGOINT signal by interrupting the operation of the data processing systems.
  • the interruption circuit 109 is a conventional circuit that is used with PDP11 data processing systems so it is not disclosed in any detail.
  • the data processing system 10 will process any such interruption by determining whether the portion of the memory 43 in FIG. 4 that is allocated to the target system is full. As previously indicated, one location corresponding to each of the data processing systems in the memory 41 contains that information. The central processor unit 13 in FIG. 1 obtains this status information by reading the information stored in that storage location during a bus reading operation.
  • the bus writing operation that produces the request to simultaneously load the address of the storage location into the ADDRESS latch 52 in FIG. 3. If that is not done, then the reading operation must be preceeded by another bus writing operation.
  • the central processor unit 13 transmits onto the data conductors 14D signals corresponding to the memory address and the C REQ signal.
  • the appearance of the C REQ signal during the writing operation maintains the stage 53 in a set state to retain control over the interconnection unit 17. In either case, so long as the CB RDY signal from stage 54 is asserted, gates 113 couple the address signals over address conductors 35A to an address port of the memory 41.
  • the data processing system 10 When the data processing system 10 performs a bus reading operation, it addresses the control data buffer 41. Also the BC1 signal is inactive, so the AND gate 84 in FIG. 6 is not energized and the DATO signal is inactive. Thus, the differential transmitters 85 are inactive while the data transceivers 88 are conditioned to convey signals from differential receivers 112 onto the conductors 14D upon receipt of the BSSYN signal.
  • the A1 and A2 address signals from the differential receivers 89A in FIG. 5 now condition the decoder 90 so that it transmits a CMMSYN signal upon receiving BMSYN signal from the differential receivers 91.
  • the decoder 90 can not transmit either the LCCSR or LDCSR signals that load information into the control status registers 40 and 42 respectively.
  • an AND gate 115 is energized during either reading or writing operations by the BMSYN signal after the delay determined by the delay circuit 94. In this case, the decoder 90 transmits a delayed CMMSYN signal that is transmitted onto a corresponding one of the conductors 35A.
  • This CMMSYN signal then is conveyed to a memory control circuit 116 in the control section 34 of FIG. 3.
  • the data in the addressed location is retrieved and produced at an output port of the memory 41 and then conveyed through transceivers 117 onto conductors 35D as CBD signals.
  • the memory control circuit 116 transmits a CMSSYN signal onto the bus section 35A.
  • This signal is conveyed to the OR gate 97 in FIG. 5.
  • the OR gate 97 When the OR gate 97 is energized, it transmits the MUXEN signal that enables the multiplexer 47 to couple the signals from the CBD conductors 35D to the bus 30.
  • the OR gates 97 and 98 and the AND gate 99 cause the differential transmitter 100 to generate the SS signal thereby to terminate the bus reading operation.
  • the central processor 13 releases its control over the interconnection unit 17. This is done by performing another bus writing operation to the control status register 40. This writing operation, however, clears the C REQ stage 53 in FIG. 3. As a result, the flip-flop 53 is cleared and the AND gate 105 is deenergized. This shifts the C REQ signal from the AND gate 105 to a non-asserted state, so the next T2 timing pulse from the clock 103 clears the flip-flop 106 thereby terminating the CBGRANT signal. When the CBGRANT signal shifts to a non-asserted level, it enables the counter 103 to respond to T1 pulses from the clock 102. Thus, the counter 103 identifies successive data processing systems until it identifies a system in which the C REQ signal is active whereupon control is passed to that data processing system.
  • the next step in transferring data to the memory 43 in FIG. 4 is to obtain exclusive control over the data section 36.
  • the central procesor unit 13, or other master unit obtains control by performing a bus writing operation to the data section control status register 42.
  • the signals that are transmitted in conjunction with this operation are conveyed through the host processor interface 20 in FIG. 6 and are received by the circuitry in FIG. 5.
  • the A1 and A2 signals now identify the register 42, so the decoder 90 transmits the LDCSR pulse that is applied to the circuitry corresponding to the write-only and read-write stages of the register 42.
  • the data transmitted from the central processor unit 13 will include a signal for setting the W/R stage 64 thereby to indicate that data will be written into the memory 43, a signal for setting the DM REQ stage 67 and a signal for setting the D IE stage 70.
  • the identification number for the target data processing system will appear in the destination stages 66.
  • the data section 36 also contains a counter 132 that is enabled so long as a D GRANT flip-flop 133 is cleared.
  • Each T1 signal from the clock 192 shown in FIG. 3 advances the counter.
  • a comparator such as the comparator 131 in FIG. 4 energizes the AND gate 130 if the flip-flop 67 is set.
  • the AND gate 130 transmits a D REQ signal onto conductors 37C. This signal is received by the flip-flop 133 which is then set on the subsequent T2 pulse thereby to prevent any further advance of the counter 132.
  • the D GRANT signal from the flip-flop 133 is then coupled back over the conductors 37C to energize the AND gate 68 that is enabled by the flip-flop 67 and the comparator 131 to transmit the DM RDY signal.
  • the flip-flop 70 in FIG. 4 If the flip-flop 70 in FIG. 4 is set, it indicates that interrupts are enabled.
  • the DM RDY signal energizes an OR gate 134.
  • the NAND gate 71C enables the AND gate 135 to be energized so that an AND gate 136, also shown in FIG. 5, causes DG signals to be transmitted onto the conductors 30.
  • the differential receivers 101 in FIG. 6 receive the DG signals and transmit a DGOINT signal that is applied to the interruption circuit 109 thereby to interrupt the data processing system 10.
  • gates 142 couple the destination code onto conductors 37A. These signals are then coupled into high-order bit positions of the address (ADR) port of the memory 43.
  • the counter 62 provides the remaining address bits. Thus, if the memory 43 has 256 storage locations allocated to each data processing system, the counter 62 will be an eight-bit counter.
  • this bus writing operation also will be used to assure that the counter 62 in FIG. 4 is cleared. This is accomplished by setting the FICLR stage represented by an AND gate 61. When this gate is energized, another AND gate 61A that is enabled by the DM RDY signal from the AND gate 68 is energized thereby producing a FIFOCLR signal that is conveyed to a clearing input of the counter 62.
  • the requesting data processing system then transfers data words in succession into the memory 43. During each such transfer, the master unit transmits the address of the fictitious data buffer 43 onto the address conductors 14A.
  • the host processor interface circuit 20 also produces the ME signal that is coupled to the bus control section over the bus 30.
  • the combination of the BMSYN signal from the differential receivers 91 and the A1 and A2 signals from the differential receivers 89A condition the decoder 90 to transmit a DMMSYN pulse.
  • each DMMSYN signal energizes an AND gate 143 if the BC1 signal is asserted.
  • An OR gate 144 and an AND gate 145 that is enabled by the DM RDY signal from the AND gate 66, transmit an FIFOCLK pulse which is applied to the counter 62 in the data section 36.
  • the counter 62 is incremented to identify the next location.
  • the C1 signal derived from the BC1 signal on the bus 14 energizes an AND gate 146 and this signal is coupled to a memory control 147 in the data section 36.
  • the memory control 147 conditions the memory 43 to perform a writing operation and conditions transceivers 148 to couple data from data conductors 37D into the data port of the memory 43.
  • the memory control 147 transmits a DSSYN signal that energizes the OR gate 97 shown in FIG. 5 thereby to enable the differential transmitters 100 to transmit the SS signal that is sent back to the data processing system as the slave synchronizing signal. This succession of steps continues until all the words to be transferred, normally the words in one sector, are loaded into the memory 43.
  • the requesting data processing system When this transfer of data into the memory 43 is complete, the requesting data processing system normally will be programmed to interrupt the target data processing system 11. This is accomplished when the requesting data processing system performs another bus writing operation to the control status register 42 shown in FIG. 2.
  • the writing operation normally will involve the transfer of signals only into the D INT, D IE and DM REQ stages, as the information in the description stages 66 in FIG. 4 usually identifies the target system.
  • the gates 142 transfer the destination onto the conductors 37A.
  • the AND gate 65 transmits the DGOINT signal onto conductors 37A.
  • each interface controller shown in FIG. 4 receives the address and DGOINT signals.
  • Each data section additionally contains a data address comparator, such as the data address comparator 150 shown in FIG. 4, that responds to the receipt of its own address from the conductors 37A.
  • the data address comparator 150 enables a NAND gate 151 so that the trailing edge of the DGOINT signal from the bus 37A sets the DI PEN flip-flop 72.
  • the resulting DI PEN signal energizes the OR gate 134 so the AND gates 135 and 136 produce the DG signal. As previously described, this interrupts the system connected to the interface circuit, in this case the target data processing system. Therefore, this circuitry enables a cross interruption to be performed.
  • the requesting data processing system releases the data section 36. This is accomplished by performing another bus writing operation to the control status register 42, this time, however, clearing the DM REQ stage 67, so the flip-flop 67 is cleared and the AND gate 130 is de-energized. As a result the next T2 pulse clears the D GRANT flip-flop 133 and enables the counter 132 to advance to the next number in sequence. Clearing the DGRANT flip-flop 133 also disables the AND gate 68 thereby terminating the DM RDY signal and all the gates that transfer data onto the conductors 37.
  • the target data processing system processes an interruption routine which performs a number of operations. Initially, the target data processing system receives certain control information from the memory 41. This is accomplished when the target central processor unit performs a bus writing operation to gain control over the control section 34 as previously described. Once control is granted to the target data processing system, it can read appropriate ones of the storage locations in its memory correponding to the memory 41 in FIG. 3 and then can interpret that information. More specifically, the central processor unit performs a bus writing operation to load the address of the control status register 42 into the address latches corresponding to the latches 52 shown in FIG. 3 and to keep the latch 53 set.
  • the target data processing system performs a bus reading operation identifying the control section data buffer 41 on the address conductors 14A.
  • the decoder 90 in FIG. 5 produces a CMMSYN pulse in response to the BMSYN signal from the differential receivers 91 and A1 and A2 address signals from the differential receivers 89A.
  • the CMMSYN pulse is coupled onto conductors 35A and then to the memory control 116 thereby to couple the output data from the memory 41 onto the conductors 35D through the transceivers 117.
  • the AND gate 92 in FIG. 5 is de-energized while an AND gate 153 transmits a DREN signal.
  • the transceivers 45 and 46 are conditioned to transfer data from the memories 41 and 43 back onto the bus 30.
  • the CB RDY signal is active so these signals are coupled through the transceiver 45 to the multiplexer 47.
  • the multiplexer 47 conveys signals to the differential transmitters 48 that are energized in response to the DREN signal from the AND gate 153 thereby to transfer the data from the corresponding storage location in the memory 41 onto the conductors 30.
  • the target central processor unit performs another bus writing operation to clear the stage 53 and thereby release its control over the interconnection unit 17.
  • the target data processing system obtains control of the data section 36 shown in FIG. 4 as previously described. Then the system performs a sequence of bus reading operations from the fictitious data buffer 42. During each successive reading operation, the decoder 90 in FIG. 5, transmits the DMMSYN pulse. In addition the AND gate 153 is energized while the AND gate 92 is not energized.
  • the DMMSYN signal energizes an AND gate 152 enabled because the WR stage 64 is cleared and the BC1 signal is not asserted during a reading operation. This enables the OR gate 144 and AND gate 145 to transmit the FIFOCLK pulses to increment the counter 62 thereby to identify successive locations.
  • the high-order bits designating the target system will be transmitted into the data address latches 66.
  • the data passes from the memory 43 and transceivers 148 onto conductors 37D. As shown in FIG. 5, these DBD signals are received by transceivers 46 that are conditioned to couple data into the multiplexer 47. As the data buffer is addressed, the multiplexer 47 is conditioned to couple the DD signals from the transceivers 46 to the differential transmitters 48, these transmitters having been turned on by the DREN signal from the AND gate 153.
  • the target data processing system performs another bus writing operation thereby to clear the DM REQ stage 67 and terminates its exclusive control over the data section of the interconnection unit 17.
  • the target data processing system again requests control over the control section 34 to clear any bits that indicate that its buffer is full thereby to allow other data processing systems to transfer files to it. It might also transfer information into a location in the memory 41 corresponding to the requesting central processor unit in order to positively acknowledge its receipt of the file. If such an exchange occurs, then a data writing operation to the control section data buffer 41 occurs. After this information is stored in the memory 41 the target system performs another bus writing operation to its control status register 40 to transfer to the latch 52 the number of the requesting system and to transfer a C INT signal to the C INT stage 51.
  • a NAND gate corresponding to the NAND gate 51 transmits the CGOINT signal that is received by an AND gate 160.
  • each interface control section contains a control address comparator 161.
  • the requesting data processing system will receive the address from the target system and energize the AND gate 160 thereby setting the latch 73.
  • the CI PEN signal from the latch 73 then energizes the OR gate 134 and produces the DG signals that interrupt the requesting data processing system.
  • the DI PEN latch 72 in FIG. 4 and the CI PEN latch 73 in FIG. 3 are cleared after the interruption sequences have been performed.
  • the data processing system that receives the interrupt performs a bus writing operation to address the control status register 42.
  • the data that is transmitted to this register sets the CLR INT stage 71 and the CI PEN stage 73.
  • the AND gate 71B resets the latch 73.
  • the a bus writing operation to the control status register 42 which sets the CLR INT stage 71 and the DI PEN stage 72 causes the AND gate 71A to clear the latch 72.
  • an interconnection unit which enables a number of independent data processing systems to operate together as a distributed data processing network and in which the individual data processing systems can be located at physically remote locations.
  • the interconnection unit has two basic sections: a control section and a data section. Each section connects to all the data processing systems in the network and each data processing network can access the data and control sections independently through bus writing operations. Counters in the controller grant exclusive control of the data and control sections to one requesting data processing system at a time. While a data processing system has control over a corresponding section, it can transfer information into or retrieve information from storage locations in the controller circuit that are assigned to it or to any other data processing system in the network. In addition, each data processing system can interrupt the operations of any other data processing system. These controls therefore enable the operation of each data processing independently while facilitating the transfer of files between those data processing systems as necessary.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Bus Control (AREA)
  • Multi Processors (AREA)
  • Communication Control (AREA)
  • Small-Scale Networks (AREA)
US05/865,116 1977-12-28 1977-12-28 Interconnection unit for multiple data processing systems Expired - Lifetime US4204251A (en)

Priority Applications (5)

Application Number Priority Date Filing Date Title
US05/865,116 US4204251A (en) 1977-12-28 1977-12-28 Interconnection unit for multiple data processing systems
GB7850070A GB2011680B (en) 1977-12-28 1978-12-28 Interconnection unit for multiple data processing systems
GB8139088A GB2091917B (en) 1977-12-28 1978-12-28 Interconnection unit for multiple data processing systems
JP16459178A JPS54102932A (en) 1977-12-28 1978-12-28 Cross connecting unit for multiple data processing system
DE19782856483 DE2856483A1 (de) 1977-12-28 1978-12-28 Verbindungseinheit fuer datenverarbeitungssysteme

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US05/865,116 US4204251A (en) 1977-12-28 1977-12-28 Interconnection unit for multiple data processing systems

Publications (1)

Publication Number Publication Date
US4204251A true US4204251A (en) 1980-05-20

Family

ID=25344757

Family Applications (1)

Application Number Title Priority Date Filing Date
US05/865,116 Expired - Lifetime US4204251A (en) 1977-12-28 1977-12-28 Interconnection unit for multiple data processing systems

Country Status (4)

Country Link
US (1) US4204251A (de)
JP (1) JPS54102932A (de)
DE (1) DE2856483A1 (de)
GB (2) GB2011680B (de)

Cited By (45)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1981002477A1 (en) * 1980-02-28 1981-09-03 Intel Corp Data processing system
US4388686A (en) * 1980-10-20 1983-06-14 General Electric Company Communication system for distributed control arrangement
US4390943A (en) * 1979-12-26 1983-06-28 Honeywell Information Systems Inc. Interface apparatus for data transfer through an input/output multiplexer from plural CPU subsystems to peripheral subsystems
US4402046A (en) * 1978-12-21 1983-08-30 Intel Corporation Interprocessor communication system
US4423482A (en) * 1981-06-01 1983-12-27 Sperry Corporation FIFO Register with independent clocking means
US4443846A (en) * 1980-12-29 1984-04-17 Sperry Corporation Dual port exchange memory between multiple microprocessors
US4449182A (en) * 1981-10-05 1984-05-15 Digital Equipment Corporation Interface between a pair of processors, such as host and peripheral-controlling processors in data processing systems
US4453211A (en) * 1981-04-28 1984-06-05 Formation, Inc. System bus for an emulated multichannel system
US4484273A (en) * 1982-09-03 1984-11-20 Sequoia Systems, Inc. Modular computer system
US4484266A (en) * 1981-12-11 1984-11-20 The United States Of America As Represented By The Secretary Of The Navy Externally specified index peripheral simulation system
US4491909A (en) * 1981-03-18 1985-01-01 International Business Machines Corporation Data processing system having shared memory
US4494190A (en) * 1982-05-12 1985-01-15 Honeywell Information Systems Inc. FIFO buffer to cache memory
EP0145748A1 (de) * 1983-05-25 1985-06-26 Ramtek Corporation Verfahren und vorrichtung zur übertragung zwischen multiprozessoren
EP0174446A2 (de) * 1984-08-03 1986-03-19 International Business Machines Corporation Verteiltes Multiverarbeitungssystem
US4603383A (en) * 1980-12-05 1986-07-29 Minolta Camera Kabushiki Kaisha Apparatus for direct data transfer among central processing units
US4633432A (en) * 1981-09-11 1986-12-30 Sharp Kabushiki Kaisha System for transferring document data to non-volatile memory in a word processing apparatus
EP0240667A2 (de) * 1986-03-12 1987-10-14 Hitachi, Ltd. Prozessor
US4731724A (en) * 1984-11-23 1988-03-15 Sintra System for simultaneous transmission of data blocks or vectors between a memory and one or a number of data-processing units
US4792896A (en) * 1983-12-07 1988-12-20 516277 Ontario Limited Storage controller emulator providing transparent resource sharing in a computer system
US4797885A (en) * 1985-12-11 1989-01-10 Hitachi, Ltd. Distributed processing system and method
US4860244A (en) * 1983-11-07 1989-08-22 Digital Equipment Corporation Buffer system for input/output portion of digital data processing system
US4862350A (en) * 1984-08-03 1989-08-29 International Business Machines Corp. Architecture for a distributive microprocessing system
EP0342596A2 (de) * 1988-05-16 1989-11-23 Canon Kabushiki Kaisha Verbindungssteuerungsvorrichtung
US4885739A (en) * 1987-11-13 1989-12-05 Dsc Communications Corporation Interprocessor switching network
US4953096A (en) * 1986-08-15 1990-08-28 Hitachi, Ltd. Test method and apparatus for distributed system
US4958271A (en) * 1984-12-07 1990-09-18 Nec Corporation Transfer control equipment
EP0422310A1 (de) * 1989-10-10 1991-04-17 International Business Machines Corporation Verteilter Mechanismus für die schnelle Planung von gemeinsamen Objekten
US5084836A (en) * 1988-04-01 1992-01-28 Kokusai Denshin Denwa Co., Ltd. Parallel signal processing system
US5132973A (en) * 1989-11-06 1992-07-21 Hewlett-Packard Company Testable embedded RAM arrays for bus transaction buffering
US5146561A (en) * 1988-06-02 1992-09-08 Sears Communications Network, Inc. Communication network data manager system
US5283907A (en) * 1986-05-20 1994-02-01 Robert Bosch Gmbh Process for safeguarding transfer of data from a temporary part into a main part of a non-volatile memory
US5297260A (en) * 1986-03-12 1994-03-22 Hitachi, Ltd. Processor having a plurality of CPUS with one CPU being normally connected to common bus
US5450603A (en) * 1992-12-18 1995-09-12 Xerox Corporation SIMD architecture with transfer register or value source circuitry connected to bus
US5544347A (en) * 1990-09-24 1996-08-06 Emc Corporation Data storage system controlled remote data mirroring with respectively maintained data indices
US5574929A (en) * 1993-06-23 1996-11-12 Koninklijke Ptt Nederland N.V. Processor circuit comprising a first processor, a memory and a peripheral circuit, and system comprising the processor circuit and a second processor
US5889935A (en) * 1996-05-28 1999-03-30 Emc Corporation Disaster control features for remote data mirroring
US5892926A (en) * 1996-12-30 1999-04-06 Compaq Computer Corporation Direct media independent interface connection system for network devices
US5909052A (en) * 1986-03-12 1999-06-01 Hitachi, Ltd. Semiconductor device having plural chips with the sides of the chips in face-to-face contact with each other in the same crystal plane
US5970471A (en) * 1996-03-22 1999-10-19 Charles E. Hill & Associates, Inc. Virtual catalog and product presentation method and apparatus
US6029142A (en) * 1992-04-10 2000-02-22 Charles E. Hill & Associates, Inc. Electronic catalog system and method
US6052797A (en) * 1996-05-28 2000-04-18 Emc Corporation Remotely mirrored data storage system with a count indicative of data consistency
US20040215856A1 (en) * 2003-04-28 2004-10-28 Yves Leduc Terminal management bus
KR100842983B1 (ko) * 2001-11-19 2008-07-01 주식회사 엘지이아이 분산처리 기능이 있는 데이터 방송용 셋탑의 대가지불방법
USRE43862E1 (en) * 2000-09-27 2012-12-11 Hewlett-Packard Development Company, L.P. Palmtop computer expansion using shaped memory access
US20130212338A1 (en) * 2012-02-14 2013-08-15 Moe Konta Multicore processor

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
IT1126475B (it) * 1979-12-03 1986-05-21 Honeywell Inf Systems Apparato di comunicazione tra piu' processori
US4463417A (en) * 1980-12-29 1984-07-31 International Business Machines Corporation Modular, processor-controlled document distribution terminal
DE3176857D1 (en) * 1980-12-29 1988-09-29 Ibm Data processing apparatus including a peripheral processing complex
FR2549621B1 (fr) * 1983-07-19 1988-09-16 Telecommunications Sa Systeme multiprocesseur pour communication des processeurs entre eux
GB2138183B (en) * 1984-01-28 1986-09-24 Standard Telephones Cables Ltd Multi-procesor system
GB2177824A (en) * 1985-07-02 1987-01-28 Director Computer Products Lim Computer peripheral controller system
GB8820369D0 (en) * 1988-08-26 1988-09-28 Int Computers Ltd Data processing system
EP0376003A3 (de) * 1988-12-29 1991-12-18 International Business Machines Corporation Mehrrechnersystem mit Zwischenprozessor-Datenkommunikationseinrichtungen
GB2296797B (en) * 1994-12-13 2000-02-16 Antech Information Systems Int An interface system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3400372A (en) * 1965-02-16 1968-09-03 Ibm Terminal for a multi-data processing system
US3753234A (en) * 1972-02-25 1973-08-14 Reliance Electric Co Multicomputer system with simultaneous data interchange between computers
US3916380A (en) * 1974-11-06 1975-10-28 Nasa Multi-computer multiple data path hardware exchange system
US3940743A (en) * 1973-11-05 1976-02-24 Digital Equipment Corporation Interconnecting unit for independently operable data processing systems
US3984819A (en) * 1974-06-03 1976-10-05 Honeywell Inc. Data processing interconnection techniques

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3665412A (en) * 1970-07-20 1972-05-23 Informalique Comp Int Numerical data multi-processor system
JPS561653B2 (de) * 1975-03-03 1981-01-14
JPS51102529A (ja) * 1975-03-07 1976-09-10 Hitachi Ltd Kiokusochisenjuhoshiki

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3400372A (en) * 1965-02-16 1968-09-03 Ibm Terminal for a multi-data processing system
US3753234A (en) * 1972-02-25 1973-08-14 Reliance Electric Co Multicomputer system with simultaneous data interchange between computers
US3940743A (en) * 1973-11-05 1976-02-24 Digital Equipment Corporation Interconnecting unit for independently operable data processing systems
US3984819A (en) * 1974-06-03 1976-10-05 Honeywell Inc. Data processing interconnection techniques
US3916380A (en) * 1974-11-06 1975-10-28 Nasa Multi-computer multiple data path hardware exchange system

Cited By (68)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4402046A (en) * 1978-12-21 1983-08-30 Intel Corporation Interprocessor communication system
US4390943A (en) * 1979-12-26 1983-06-28 Honeywell Information Systems Inc. Interface apparatus for data transfer through an input/output multiplexer from plural CPU subsystems to peripheral subsystems
WO1981002477A1 (en) * 1980-02-28 1981-09-03 Intel Corp Data processing system
US4388686A (en) * 1980-10-20 1983-06-14 General Electric Company Communication system for distributed control arrangement
US4603383A (en) * 1980-12-05 1986-07-29 Minolta Camera Kabushiki Kaisha Apparatus for direct data transfer among central processing units
US4443846A (en) * 1980-12-29 1984-04-17 Sperry Corporation Dual port exchange memory between multiple microprocessors
US4491909A (en) * 1981-03-18 1985-01-01 International Business Machines Corporation Data processing system having shared memory
US4453211A (en) * 1981-04-28 1984-06-05 Formation, Inc. System bus for an emulated multichannel system
US4423482A (en) * 1981-06-01 1983-12-27 Sperry Corporation FIFO Register with independent clocking means
US4633432A (en) * 1981-09-11 1986-12-30 Sharp Kabushiki Kaisha System for transferring document data to non-volatile memory in a word processing apparatus
US4449182A (en) * 1981-10-05 1984-05-15 Digital Equipment Corporation Interface between a pair of processors, such as host and peripheral-controlling processors in data processing systems
US4484266A (en) * 1981-12-11 1984-11-20 The United States Of America As Represented By The Secretary Of The Navy Externally specified index peripheral simulation system
US4494190A (en) * 1982-05-12 1985-01-15 Honeywell Information Systems Inc. FIFO buffer to cache memory
US4484273A (en) * 1982-09-03 1984-11-20 Sequoia Systems, Inc. Modular computer system
EP0145748A4 (de) * 1983-05-25 1988-03-30 Ramtek Corp Verfahren und vorrichtung zur übertragung zwischen multiprozessoren.
EP0145748A1 (de) * 1983-05-25 1985-06-26 Ramtek Corporation Verfahren und vorrichtung zur übertragung zwischen multiprozessoren
US4860244A (en) * 1983-11-07 1989-08-22 Digital Equipment Corporation Buffer system for input/output portion of digital data processing system
US4792896A (en) * 1983-12-07 1988-12-20 516277 Ontario Limited Storage controller emulator providing transparent resource sharing in a computer system
US4862350A (en) * 1984-08-03 1989-08-29 International Business Machines Corp. Architecture for a distributive microprocessing system
EP0174446A3 (en) * 1984-08-03 1988-01-07 International Business Machines Corporation Distributed multiprocessing system
EP0174446A2 (de) * 1984-08-03 1986-03-19 International Business Machines Corporation Verteiltes Multiverarbeitungssystem
US4731724A (en) * 1984-11-23 1988-03-15 Sintra System for simultaneous transmission of data blocks or vectors between a memory and one or a number of data-processing units
US4958271A (en) * 1984-12-07 1990-09-18 Nec Corporation Transfer control equipment
US4797885A (en) * 1985-12-11 1989-01-10 Hitachi, Ltd. Distributed processing system and method
EP0240667A2 (de) * 1986-03-12 1987-10-14 Hitachi, Ltd. Prozessor
US6379998B1 (en) 1986-03-12 2002-04-30 Hitachi, Ltd. Semiconductor device and method for fabricating the same
US5968150A (en) * 1986-03-12 1999-10-19 Hitachi, Ltd. Processor element having a plurality of CPUs for use in a multiple processor system
EP0240667A3 (en) * 1986-03-12 1989-12-27 Hitachi, Ltd. Processor
US5909052A (en) * 1986-03-12 1999-06-01 Hitachi, Ltd. Semiconductor device having plural chips with the sides of the chips in face-to-face contact with each other in the same crystal plane
US5568617A (en) * 1986-03-12 1996-10-22 Hitachi, Ltd. Processor element having a plurality of processors which communicate with each other and selectively use a common bus
US5297260A (en) * 1986-03-12 1994-03-22 Hitachi, Ltd. Processor having a plurality of CPUS with one CPU being normally connected to common bus
US5283907A (en) * 1986-05-20 1994-02-01 Robert Bosch Gmbh Process for safeguarding transfer of data from a temporary part into a main part of a non-volatile memory
US4953096A (en) * 1986-08-15 1990-08-28 Hitachi, Ltd. Test method and apparatus for distributed system
US4885739A (en) * 1987-11-13 1989-12-05 Dsc Communications Corporation Interprocessor switching network
US5084836A (en) * 1988-04-01 1992-01-28 Kokusai Denshin Denwa Co., Ltd. Parallel signal processing system
EP0342596A3 (en) * 1988-05-16 1990-08-01 Canon Kabushiki Kaisha Communication control device
EP0342596A2 (de) * 1988-05-16 1989-11-23 Canon Kabushiki Kaisha Verbindungssteuerungsvorrichtung
US5455688A (en) * 1988-05-16 1995-10-03 Canon Kabushiki Kaisha Communication control device for controlling the flow of data between a plurality of devices
US5146561A (en) * 1988-06-02 1992-09-08 Sears Communications Network, Inc. Communication network data manager system
EP0422310A1 (de) * 1989-10-10 1991-04-17 International Business Machines Corporation Verteilter Mechanismus für die schnelle Planung von gemeinsamen Objekten
US5428781A (en) * 1989-10-10 1995-06-27 International Business Machines Corp. Distributed mechanism for the fast scheduling of shared objects and apparatus
US5132973A (en) * 1989-11-06 1992-07-21 Hewlett-Packard Company Testable embedded RAM arrays for bus transaction buffering
US5544347A (en) * 1990-09-24 1996-08-06 Emc Corporation Data storage system controlled remote data mirroring with respectively maintained data indices
US6131088A (en) * 1992-04-10 2000-10-10 Charles E. Hill & Associates, Inc. Electronic catalog system and method
US6029142A (en) * 1992-04-10 2000-02-22 Charles E. Hill & Associates, Inc. Electronic catalog system and method
US5450603A (en) * 1992-12-18 1995-09-12 Xerox Corporation SIMD architecture with transfer register or value source circuitry connected to bus
US20040073831A1 (en) * 1993-04-23 2004-04-15 Moshe Yanai Remote data mirroring
US7055059B2 (en) 1993-04-23 2006-05-30 Emc Corporation Remote data mirroring
US6647474B2 (en) 1993-04-23 2003-11-11 Emc Corporation Remote data mirroring system using local and remote write pending indicators
US7073090B2 (en) 1993-04-23 2006-07-04 Emc Corporation Remote data mirroring system having a remote link adapter
US6625705B2 (en) 1993-04-23 2003-09-23 Emc Corporation Remote data mirroring system having a service processor
US20030167419A1 (en) * 1993-04-23 2003-09-04 Moshe Yanai Remote data mirroring system having a remote link adapter
US5742792A (en) * 1993-04-23 1998-04-21 Emc Corporation Remote data mirroring
US6173377B1 (en) 1993-04-23 2001-01-09 Emc Corporation Remote data mirroring
US6502205B1 (en) 1993-04-23 2002-12-31 Emc Corporation Asynchronous remote data mirroring system
US5574929A (en) * 1993-06-23 1996-11-12 Koninklijke Ptt Nederland N.V. Processor circuit comprising a first processor, a memory and a peripheral circuit, and system comprising the processor circuit and a second processor
US5970471A (en) * 1996-03-22 1999-10-19 Charles E. Hill & Associates, Inc. Virtual catalog and product presentation method and apparatus
US7908173B1 (en) 1996-03-22 2011-03-15 Charles E. Hill & Associates, Inc. Virtual catalog and product presentation method and apparatus
US7908176B1 (en) 1996-03-22 2011-03-15 Charles E. Hill & Associates, Inc. Virtual catalog and product presentation method and apparatus
US6052797A (en) * 1996-05-28 2000-04-18 Emc Corporation Remotely mirrored data storage system with a count indicative of data consistency
US6044444A (en) * 1996-05-28 2000-03-28 Emc Corporation Remote data mirroring having preselection of automatic recovery or intervention required when a disruption is detected
US5901327A (en) * 1996-05-28 1999-05-04 Emc Corporation Bundling of write data from channel commands in a command chain for transmission over a data link between data storage systems for remote data mirroring
US5889935A (en) * 1996-05-28 1999-03-30 Emc Corporation Disaster control features for remote data mirroring
US5892926A (en) * 1996-12-30 1999-04-06 Compaq Computer Corporation Direct media independent interface connection system for network devices
USRE43862E1 (en) * 2000-09-27 2012-12-11 Hewlett-Packard Development Company, L.P. Palmtop computer expansion using shaped memory access
KR100842983B1 (ko) * 2001-11-19 2008-07-01 주식회사 엘지이아이 분산처리 기능이 있는 데이터 방송용 셋탑의 대가지불방법
US20040215856A1 (en) * 2003-04-28 2004-10-28 Yves Leduc Terminal management bus
US20130212338A1 (en) * 2012-02-14 2013-08-15 Moe Konta Multicore processor

Also Published As

Publication number Publication date
JPS54102932A (en) 1979-08-13
DE2856483A1 (de) 1979-07-12
DE2856483C2 (de) 1993-02-11
GB2091917A (en) 1982-08-04
GB2091917B (en) 1983-03-02
GB2011680A (en) 1979-07-11
JPS6218949B2 (de) 1987-04-25
GB2011680B (en) 1983-01-06

Similar Documents

Publication Publication Date Title
US4204251A (en) Interconnection unit for multiple data processing systems
US3940743A (en) Interconnecting unit for independently operable data processing systems
US4229791A (en) Distributed arbitration circuitry for data processing system
US4488217A (en) Data processing system with lock-unlock instruction facility
US4991079A (en) Real-time data processing system
EP0029975B1 (de) Multiprozessorsystem
US3648252A (en) Multiprogrammable, multiprocessor computer system
US4860244A (en) Buffer system for input/output portion of digital data processing system
US3673576A (en) Programmable computer-peripheral interface
US4485438A (en) High transfer rate between multi-processor units
US5093910A (en) Serial data transmission between redundant channels
US4381542A (en) System for interrupt arbitration
EP1422626B1 (de) Multi-Kern Kommunikationsmodul, Datenkommunikationssystem mit einem derartigen Modul und Datenkommunikationsverfahren
KR100192529B1 (ko) 원격버스상의 중재를 고정시키는 방법 및 장치
US5404462A (en) Dual bus interface transfer system for central processing module
US6823411B2 (en) N-way psuedo cross-bar having an arbitration feature using discrete processor local busses
JPH071495B2 (ja) データ処理システム
US5228127A (en) Clustered multiprocessor system with global controller connected to each cluster memory control unit for directing order from processor to different cluster processors
US4509115A (en) Two-port memory controller
JPH04312160A (ja) マルチプロセッサシステムおよびそのメッセージ送受信制御装置
US4417303A (en) Multi-processor data communication bus structure
US5146605A (en) Direct control facility for multiprocessor network
JPH04257054A (ja) チャネル間接続装置
JPH0198048A (ja) 周辺装置制御装置およびアダプタ・インターフェース
US4245303A (en) Memory for data processing system with command and data buffering

Legal Events

Date Code Title Description
AS Assignment

Owner name: ATEX, INC. A DE CORPORATION, MASSACHUSETTS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNOR:BRUDEVOLD, FINN;REEL/FRAME:005753/0395

Effective date: 19910611

AS Assignment

Owner name: ATEX PUBLISHING SYSTEMS CORPORATION, MASSACHUSETTS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ATEX INC.;REEL/FRAME:006983/0682

Effective date: 19940110